Present address: Clotilde Biard, Equipe Ecologie Evolutive, UMR CNRS 5561 Biogéosciences, Université de Bourgogne, 6 bd Gabriel, F-21000 Dijon, France.
An analysis of pre- and post-hatching maternal effects mediated by carotenoids in the blue tit
Article first published online: 3 AUG 2006
Journal of Evolutionary Biology
Volume 20, Issue 1, pages 326–339, January 2007
How to Cite
BIARD, C., SURAI, P. F. and MØLLER, A. P. (2007), An analysis of pre- and post-hatching maternal effects mediated by carotenoids in the blue tit. Journal of Evolutionary Biology, 20: 326–339. doi: 10.1111/j.1420-9101.2006.01194.x
- Issue published online: 18 DEC 2006
- Article first published online: 3 AUG 2006
- Received 4 April 2006; revised 9 June 2006; accepted 16 June 2006
- cell-mediated immune response;
- fledgling condition;
- maternal effects;
- plumage colour
Maternal effects increase phenotypic plasticity in offspring traits and may therefore facilitate adaptation to environmental variability. Carotenoids have been hypothesized to mediate costs of reproduction in females as well as maternal effects. However, assessing potential transgenerational and population consequences of environmental availability of carotenoids requires a better understanding of mechanisms of maternal effects mediated by these antioxidant pigments. Manipulating dietary availability of carotenoids to egg-laying female blue tits and subsequently cross-fostering nestlings between female treatments allowed us to specifically investigate the relative importance of maternal effects through egg carotenoids and through post-hatching care mediated by antioxidants in females. Nestling body size and mass and plasma antioxidants were not significantly affected by pre- or post-hatching maternal effects mediated by antioxidants, although both types of maternal effects in interaction explained the variation in growth, as measured by wing length. Development of the ability to mount a cell-mediated immune response as well as its temporal dynamics was influenced by both pre- and post-hatching maternal effects, with an advantage to nestlings originating from, or reared by, carotenoid-supplemented females. In addition, nestlings reared by carotenoid-fed females had a lower blood sedimentation rate, indicating that they may have been less infected than nestlings from controls. Finally, prehatching maternal effects in interaction with nestling plasma carotenoid levels affected the development of carotenoid-based plumage. Maternal effects mediated by carotenoids may thus act as a proximate factor in development and phenotypic plasticity in traits associated with nestling fitness, such as immune response and ability to metabolize and use antioxidants, and ultimately participate in the evolution of phenotypic traits.
The expression of a phenotype is determined by the individual's genotype, the environment in which it develops and their interaction. The relative importance of genetic and environmental components is linked to the selective advantage of a trait (Falconer & Mackay, 1996). Phenotypic traits closely associated with fitness such as life history traits tend to show less genetic variation and a stronger environmental component than traits less associated with fitness such as morphological traits (Mousseau & Roff, 1987; Price & Schluter, 1991; Christe et al., 2000; Endler, 2000). However, maternal effects can indirectly influence early developmental conditions thereby modulating the phenotype and fitness of offspring (Mousseau & Fox, 1998). Maternal effects may represent an important mechanism maintaining the additive genetic variance in natural populations because maternal effects may facilitate adaptation to the environment without causing fixation of advantageous alleles (Wade, 1998).
Birds are of particular interest for the study of early maternal effects in animals, as females have to provide eggs with all necessary nutrients for the development of the embryo before laying. In addition, parents in many species provide offspring with post-hatching care until independence (for a review of parental effects in birds, see Price, 1998). A particular maternal effect acting through egg quality and nutrient composition that has recently received much interest is carotenoid content of egg yolk (Royle et al., 2001, 2003; Blount et al., 2002; Hõrak et al., 2002; Saino et al., 2002). Carotenoids are pigments with antioxidant properties synthesized by plants, algae and fungi that animals must therefore acquire from food (Goodwin, 1984). These pigments are involved in the development of bright yellow to red coloration of feathers and skin (reviewed in Møller et al., 2000). Carotenoids are part of the antioxidant system and play important stimulating and modulating roles in the immune system (reviewed in Bendich, 1993; Chew, 1993; Møller et al., 2000). In addition, carotenoids are essential to embryonic and early post-hatching development (reviewed in Surai et al., 2001a,b), and they have been shown to influence the nestling phenotype (plumage colour: Biard et al., 2005, 2006; immune response: Saino et al., 2003; ability to use antioxidants in juvenile and adult life: Hõrak et al., 2000; Blount et al., 2003; Koutsos et al., 2003; sexual signals: McGraw et al., 2005). However, carotenoids may be a limiting resource as a consequence of environmental variability in their availability, differences in foraging ability of individuals or differences in absorption or metabolizing efficiency among individuals (reviewed in Olson & Owens, 1998; Møller et al., 2000). There is indeed growing evidence that carotenoid availability is limiting for wild populations both the outside (Slagsvold & Lifjeld, 1985; Hill et al., 2002) and during the reproductive season (Blount et al., 2004). In addition, maternal deposition of yolk carotenoids has been shown to vary with laying order (Royle et al., 2001, 2003; Saino et al., 2002) and to depend on habitat (Hõrak et al., 2002), female condition and male attractiveness (Saino et al., 2002).
As carotenoids are essential both for self-maintenance and reproduction, they may mediate costs of reproduction in females (Blount et al., 2000). The availability of carotenoids during egg laying and its effects on egg yolk composition, female condition and offspring performance have recently been investigated in wild birds by experimentally increasing female access to carotenoids (Blount et al., 2002, 2004; Biard et al., 2005). However, such indirect manipulation of egg composition makes it difficult to disentangle the effects of carotenoid availability for the development of nestling phenotype through yolk composition from the effects of carotenoid availability on the female condition. One way to experimentally separate pre- and post-hatching maternal effects is to use a cross-fostering design, in which nestlings are raised either by their genetic mother or a foster mother. This approach is used to separate genetic and prehatching maternal effects from common rearing environment effects, including the post-hatching maternal effects (Merilä, 1996; Roff, 1998; Krist & Remeš, 2004). However, when investigating the effects of egg size or quality, this approach does not control for positive covariation between direct genetic and indirect maternal effects, and only partially controls for covariation between maternal and environmental effects (Krist & Remeš, 2004). Here, we present a new approach for studying maternal effects using a cross-fostering experiment, where the environmental component of the pre- and post-hatching maternal effects of interest is experimentally manipulated, allowing to directly assess their relative role for the development of nestling phenotype by comparing treated and control groups while the covariance among siblings because of common origin or rearing environments is controlled for.
We previously reported the effects of an increase in carotenoid availability during laying on reproduction in the blue tit (Parus caeruleus L.) (1758) (Biard et al., 2005). Carotenoid supplementation of egg-laying females resulted in a significant increase in carotenoid concentration in egg yolk after the treatment. Nestlings from eggs laid by carotenoid-supplemented females had longer tarsi, more circulating leucocytes and brighter carotenoid-based plumage colour than nestlings from control females, but nestlings from the two groups did not differ significantly in body mass or plasma antioxidants (Biard et al., 2005). In that experiment, it is likely that prehatching maternal effects occurred through modification of egg composition, but there may also have been post-hatching maternal effects. Increasing the availability of carotenoids during laying had a long-term beneficial effect on female antioxidant function at the time of chick-rearing, although females from the two treatment groups did not differ significantly in body condition (Biard et al., 2005). Females from both treatments thus differed in their need for dietary antioxidants and/or in their ability to fulfil energetically demanding activities that may increase their level of oxidative stress. As a consequence, females may provide their nestlings with developmental environments of different quality.
In order to investigate the relative importance of pre- and post-hatching maternal effects mediated by carotenoids, we performed a cross-fostering experiment on blue tit nestlings following dietary supplementation of egg laying females with carotenoids. The cross-fostering took place within pairs of nests from different female treatments in such a way that a carotenoid-supplemented female raised half her own brood along with half the brood of a paired control female, and vice versa. Nestling body size and condition, cell-mediated immune response and carotenoid-based plumage colour were investigated in relation to natal and rearing female treatment, and covariance among siblings because of common origin or common rearing environment was estimated. Previous cross-fostering experiments in passerines have shown that tarsus length and body condition in fledglings are mostly determined by common origin, although the amount of genetic variation expressed depends on the environmental conditions (e.g. good or bad year, brood size) (Gebhardt-Heinrich & van Noordwijk, 1991; Thessing & Ekman, 1994; Smith & Wettermark, 1995; Merilä, 1996, 1997; Meriläet al., 1999; Kruuk et al., 2003). Given the importance of body condition at fledging for over-winter survival and recruitment in Parus species (Tinbergen & Boerlijst, 1990; Naef-Daenzer et al., 2001), we would expect a stronger effect of rearing environment compared with the common origin effect on fledgling body mass. There is evidence for a strong genetic basis of immune function, either for resistance to parasites or development of cell-mediated immune response (Møller, 1990; Saino et al., 1997), but environmental and maternal effects might also be important (Brinkhof et al., 1999; Soler et al., 2003; Cichoñet al., 2006; for a colonial species, see Christe et al., 2000). To our knowledge, only one study has investigated the relative importance of genetic and environmental components of nestling plasma carotenoids in the wild, finding a low genetic determinism, while rearing environment explained 61 % of the variance (Bortolotti et al., 2000). Finally, previous cross-fostering experiments on nestling carotenoid-based plumage colour reported both common origin and common rearing effects, although the latter seemed to be the main factor (Slagsvold & Lifjeld, 1985; Hõrak et al., 2000; Fitze et al., 2003; Tschirren et al., 2003). Any difference in nestling characteristics according to natal female treatment would reveal prehatching maternal effects, while any effect of rearing female treatment would reveal modifications in growth conditions experienced by nestlings as a result of variation in carotenoid availability to egg-laying females.
Materials and methods
Data collection and experimental protocol
Data were collected in 2002 in a population of blue tits breeding in eastern France (48°17′N, 4°18′E). The study area (about 250 ha) contained 400 nestboxes evenly distributed among an old homogenous deciduous woodland composed mainly of oak (Quercus spp.), hornbeam (Carpinus betulus) and beech (Fagus sylvatica). Nests were regularly inspected for laying date, clutch size and incubation date, hatching date, and number of hatchlings and fledglings.
The complete experimental protocol is explained in detail in Biard et al. (2005). Females were provided with supplemental food units made of either pure solid vegetable oil (control treatment) or the same oil melted with carotenoids (v/v) (20 % lutein and 0.86 % zeaxanthin; Kemin Foods FloraGlo, Nantes, France) to a final concentration of xanthophylls of 10 % (carotenoid treatment). Nests were randomly attributed to a treatment group on the day the first egg was laid (n = 79 nests) and food supplementation was stopped at clutch completion. Once incubation started, nests were not visited until the day before the estimated hatching date, and then visited each day until hatching (day 0, n = 69 successfully hatched broods).
Cross fostering of hatchlings took place between nests of control- and carotenoid-fed females matched for hatching date, brood size and mean hatchling mass. When at least two nests from different treatments and of similar clutch size hatched on the same day, their hatchlings were weighed on day 1 to the nearest 0.1 g either with an electronic field balance or with a 10 g Pesola spring balance (both methods provided identical results). A mean hatchling mass was then calculated for the nest (mean ± SE hatchling mass day 1 = 1.54 ± 0.03 g, n = 359 nestlings; 46 nests). Nests with the same hatching date were then compared, and pairs made between nests of similar brood size (maximum difference allowed: two nestlings or only one nestling if one or two eggs were still to hatch) and similar mean nestling mass (max. difference allowed: 0.2 g). Hatchlings from nests for which no pair was formed on day 1 were weighed on day 2 (mean ± SE hatchling mass day 2 = 2.11 ± 0.05 g, n = 122 nestlings; 16 nests) and compared with newly hatched nestlings. If no matching nest was found on the second day after hatching, the nest was excluded from the cross-fostering experiment (n = 35). Cross fostering nestlings between nests could only be performed when the frequency of hatching was high enough for matching nests. This induced a difference in mean and variance of hatching date between cross-fostered (117.53 ± 0.50, range: 113–123) and noncross-fostered nests (120.44 ± 0.93, range: 111–135, 110 = April, 20th) (homoscedasticity test: F[33,33] = 3.47, P = 0.0006; Satterthwaite t-test for unequal variances: d.f. = 50.6, t = 2.74, P = 0.008). Within pairs of nests, one half of the nestlings from each clutch was randomly chosen to remain in the nest and the other half was put in a warm box and quickly (<10 min) moved to the paired nest. Exactly the same number of nestlings (4 or 5) was removed/replaced between nests, to keep brood size unchanged. To allow the identification of resident and foster nestlings until a numbered aluminium ring could be fitted (usually on day 6 or 7), foster hatchlings were marked on the back and on the downy feathers with nontoxic red dye, or with a spot of dark nail varnish on the claws. The type of marking used was randomly attributed with respect to their genetic mother treatment. As the markings often quickly faded away and therefore did not allow clear identification of nestling origin, we checked for identification and, if needed, corrected nest assignment of each nestling using genetic markers (microsatellites) analysed for another study. In the possibility where coloured markings had modified the parent's feeding behaviour (e.g. Götmark & Ahlström, 1997), it would only have been for a very short period of time, and would have contributed to noise in the data, rather than have confounded the results with respect to maternal treatment effects. A total of 34 nests were paired, of which two pairs of nests and two nests from different pairs and treatments failed, leaving 13 pairs and two half pairs of nests for analysis.
Chicks were captured in the nest a few days before fledging, when 11–18 days old (mean ± SE: 14.9 ± 0.4 days, n = 28 nests and 240 nestlings). Because of logistic constraints in the field, nestling age at capture was positively correlated with hatching date (r = 0.86, P < 0.0001). We measured tarsus length to the nearest 0.1 mm with a calliper, wing length to the nearest 1 mm with a ruler and weighed nestlings to the nearest 0.25 g with a Pesola spring balance. A sample of 5–8 yellow feathers was plucked from the centre of the yellow breast for each bird, and stored in individual plastic bags in the dark until later colour analysis. A blood sample (50–100 μL) was taken from the brachial vein in heparinized micro-haematocrit tubes for 239 nestlings. Blood samples were stored in a cooling bag in the field, and when back in the lab, processed to measure sedimentation rate and haematocrit (Biard et al., 2005). Plasma was then separated from blood cells and stored at −20 °C. Elevated blood sedimentation rate is indicative of acute infections and inflammatory diseases, as this measure reflects levels of circulating immunoglobulins and fibrinogen (e.g. Sturkie, 1986; Svensson & Merilä, 1998).
After capture and measurements, cross-fostered nestlings were tested for their ability to raise a cell-mediated immune response to phytohaemagglutinin (PHA, lectin from Phaseolus vulgaris), which involves the combined responses of T cells, cytokines and inflammatory cells (Sharma, 1990; Davison et al., 1996; Parmentier et al., 1998). Nestlings were injected s.c. with 0.2-mg PHA (L-8754; Sigma, Saint-Quentin Fallavier, France) in 0.04-mL sterile phosphate buffered saline (PBS) within the wing patagium first plucked from feathers and marked for injection point with permanent ink. No control injection of PBS was made in the other wing web (Smits et al., 1999). Wing web thickness was measured with a spessimeter (Alpa S.p.A.; Milan, Italy) to 0.01 mm just before injection and at least 12 h after injection to assess the intensity of the immune response (mean time ± SE = 19.85 ± 0.4 h, min = 12.23 h, max = 33.5 h, n = 178). Sample size was reduced due to early fledging before recapture. The immune response index (wing web swelling) was calculated as the difference between thickness after and before injection.
Antioxidant analysis of plasma
Plasma samples (20 μL) were first mixed with 40 μL ethanol, followed by antioxidants being extracted twice with 500 μL hexane. Hexane extracts were pooled and evaporated at 60–65 °C under nitrogen flow and the residue was then dissolved in 0.1 mL dichloromethane and 0.1 mL methanol. Carotenoid composition and concentration, and vitamin A and E concentration were determined using reverse phase high performance liquid chromatography following previously published procedures (Surai, 2000; Surai et al., 2001c; for details see Biard et al., 2005). All nestling plasma samples of at least 20 μL (n = 209) were analysed for total carotenoid concentration, and a random sub-sample was analysed for vitamin E (n = 157). Samples were analysed blindly with respect to the origin of individuals. Vitamin E calculated as the summed concentrations of δ-, γ-, α-tocopherol is used in subsequent analyses. However, results obtained considering only the major antioxidant form, α-tocopherol (Surai, 2002), were qualitatively similar. Concentrations are given in μg mL−1.
Feather colour analysis
The yellow colour of the blue tit's juvenile and adult breast plumage is carotenoid-based and determined by two xanthophyll pigments, lutein and zeaxanthin (Partali et al., 1987). Breast feather colour was analysed for all 240 nestlings in the laboratory, using a spectroradiometer (Ocean Optics, Duiven, the Netherlands) (Saino et al., 1999; Hõrak et al., 2000; Biard et al., 2005), blindly with respect to the origin of samples. Feathers were illuminated at an angle of 90° with a deuterium-halogen lamp, and reflected light was measured at an angle of 45°. We calculated two colour parameters for breast feather colour: brightness in the yellow part of the spectrum was calculated as the integral of reflectance in the interval [550–625] nm (Saks et al., 2003b), and hue was calculated as an angle value using the segment classification (0°: red −90°: yellow) (Endler, 1990). Two feathers were analysed for each individual, with four measures per feather. Repeatability of measurements (intra-class correlation coefficient, Lessells & Boag, 1987) was always highly significant (all P < 0.0001) (Biard et al., 2005). Average values for the eight measurements per individual were used in subsequent statistical analyses.
All statistical analyses were made using sas v. 8.2 (SAS Institute Inc. 1999–2001, Cary, NC, USA). Tests of residuals for normality and homoscedasticity were used to check the validity of the model. Tarsus length, haematocrit and time between measures of initial and post-reaction wing web thickness were entered as covariates in all models with body mass, sedimentation and cell-mediated immune response, respectively, as dependent variables (Freckelton, 2002). In addition, initial models always included nestling age to account for seasonal variation and nestling growth, and brood size of rearing nest to account for sibling competition. Interactions and main effects were sequentially dropped from the model when not significant.
Effect of female treatment on fledgling characteristics was tested using mixed linear models (Goldstein, 2003). In order to quantify the respective parts of variance associated with genetic and prenatal parental effects, and rearing environment and post-natal parental effects, the random part of the model included both the natal and rearing nests. Natal nest models the covariance of siblings raised in different nests whereas rearing nest models the covariance of both related and unrelated siblings raised in the same nest. In addition, the random structure of the model initially included the pair of cross-fostered nests, with natal and rearing nest effects nested within pair. Including the pair of cross-fostered nest should control for common characteristics of the pair of nests (e.g. Merilä, 1996, 1997), i.e. hatching date, clutch size and mean hatchling body mass. If this random structure did not significantly improve the model's fit, then the pair covariance parameter was dropped from the model. The fixed part of the model included maternal treatment (natal female treatment) and foster female treatment (rearing female treatment) as well as their interaction with covariates. The mixed procedure was used to fit these models, using a Variance Components covariance structure and REML (restricted/residual maximum likelihood) estimation method. The Wald's test of the covariance parameter was used to assess whether the variance was significantly structured by random effect(s). The null model likelihood ratio chi-squared test was used to assess whether the model with random effect(s) provided a significantly better fit than the same model without random effect(s). If that was not the case, the model was constructed without random effect. Denominator degrees of freedom for Type 3 tests of fixed effects were estimated using the Kenward–Roger method (Spilke et al., 2005). Models were compared with Akaike's Information Criterion (AIC), and the most parsimonious model was retained (lowest AIC, Burnham & Anderson, 1998).
Significant models were followed by multiple comparisons of least square means between treatments with adjusted values of P, using the Tukey–Kramer method. Values are reported as mean ± SE.
Variance in nestling tarsus length was mostly structured by common origin, but not by common rearing environment: nestlings from the same natal nest were similar in tarsus length (covariance parameter estimate ± SE: 0.108 ± 0.036, Wald's Z = 2.98, P < 0.001) but not nestlings from the same rearing nest (covariance parameter estimate ± SE: 0.003 ± 0.010, Wald's Z = 0.30, P = 0.38) (likelihood ratio test: = 57.94, P < 0.0001). This suggests that tarsus length in this population and year was more influenced by genotype and prehatching parental effects than by rearing environment and post-hatching parental effects. However, there was no detectable effect of natal (F[1,26.3] = 0.21, P = 0.65, carotenoid group: 18.99 ± 0.05 mm, control group: 19.06 ± 0.05 mm) or rearing female treatment (F[1,12.7] = 3.24, P = 0.09, carotenoid group: 19.09 ± 0.05 mm, control group: 18.97 ± 0.05 mm) on nestling tarsus length.
The influence of common rearing environment on variance in body mass (rearing nest covariance parameter estimate ± SE: 0.248 ± 0.093, Wald's Z = 2.68, P < 0.01) was twice that of common origin (natal nest covariance parameter estimate ± SE: 0.110 ± 0.051, Wald's Z = 2.16, P < 0.05) (likelihood ratio test: = 136.32, P < 0.0001). This is indicative of a significant genetic and prehatching parental effect on mass gain, although this effect was less important than the effect of environmental conditions. This is consistent with our prediction of a larger environmental component for body mass, as expected for a trait strongly associated with fitness. However, nestling body mass did not differ with respect to natal (F[1,14.7] = 0.14, P = 0.71, carotenoid group: 10.96 ± 0.08 g, control group: 11.06 ± 0.08 g) or rearing female treatment (F[1,18.1] = 0.00, P = 0.99, carotenoid group: 11.07 ± 0.07 g, control group: 10.95 ± 0.1 g). Nestling body mass increased with nestling tarsus length (F[1,221] = 90.30, P < 0.0001) and decreased with brood size of rearing nest (F[1,22.2] = 4.59, P < 0.05).
In the same way as for body mass, there was significant genetic and prehatching parental effect on growth, although this effect was less important than the effect of environmental conditions: the influence of common rearing environment on variance in wing length was about twice that of common origin (Table 1). In a model controlling for nestling age, there was a strong effect of the interaction between natal and rearing female treatment on nestling wing length, indicating an effect of carotenoid availability through both pre- and post-hatching maternal effects on nestling growth. However, the effect of the interaction between natal and rearing female treatment was not reflected in pairwise comparisons of least square means of wing length (Table 1). Therefore, the direction of the differences in wing length between the four groups of nestlings from different natal and rearing female treatments could not be detected with our sample size.
|Covariance parameter estimates||Likelihood ratio test: = 107.47, P < 0.0001|
|Parameter||Estimate ± SE||Wald's Z||P|
|Natal nest||1.67 ± 0.80||2.08||0.02|
|Rearing nest||3.38 ± 1.32||2.56||0.005|
|Residual||3.57 ± 0.36||9.85||<0.0001|
|Type 3 tests of fixed effects|
|Effect||Estimate ± SE||Fd.f.||P|
|Tarsus length||1.85 ± 0.40||17.20 1,219||<0.0001|
|Nestling age||2.14 ± 0.22||95.12 1,30.7||<0.0001|
|Rearing female treatment**||23.62 ± 10.33*||4.92 1,211||0.03|
|Natal female treatment||0.44 ± 0.59*||0.31 1,14.2||0.59|
|Tarsus length × rearing female treatment||−1.18 ± 0.54*||4.77 1,210||0.03|
|Rearing female treatment × natal female treatment***||−1.48 ± 0.51||8.50 1,202||0.004|
Blood sedimentation rate
Nestlings were more similar for blood sedimentation rate within than among pairs of cross-fostered nests, and natal or rearing nest did not further explain covariance among nestlings (Table 2). This suggests that factors related to common characteristics of the pair of nests, such as hatching date, clutch size or mean hatchling body mass are more important in determining variation in nestling blood sedimentation rate than genetic and prehatching parental and environmental effects or immediate rearing environment. Nestlings reared by a carotenoid-fed female had a significantly smaller blood sedimentation rate (mean ± SE: 65.4 ± 0.6 %) than nestlings raised by control-fed females (66.8 ± 0.7 %). Nestlings reared by carotenoid-fed females thus circulated lower levels of immune compounds associated with infections and inflammatory diseases, such as immunoglobulins and fibrinogen. However, there was no significant effect of natal female treatment (prehatching environment) on nestling sedimentation rate (carotenoid: 66.5 ± 0.7 %; control: 65.8 ± 0.1 %) in a model accounting for nestling haematocrit and nestling age (Table 2).
|Covariance parameter estimates||Likelihood ratio test: = 47.96, P < 0.0001|
|Parameter||Estimate ± SE||Wald's Z||P|
|Pair of cross-fostered nests||0.00103 ± 0.0006||1.85||0.03|
|Natal nest (Pair)||0.000210 ± 0.0002||1.02||0.15|
|Rearing nest (Pair)||0.00007 ± 0.0002||0.42||0.34|
|Residual||0.00262 ± 0.0003||9.94||<0.0001|
|Type 3 tests of fixed effects|
|Effect||Estimate ± SE||Fd.f.||P|
|Haematocrit||0.81 ± 0.83||93.88 1,156||<0.0001|
|Nestling age||−0.013 ± 0.004||10.20 1,28.6||0.003|
|Rearing female treatment||−0.026 ± 0.008*||11.20 1,11.3||0.006|
|Natal female treatment||0.010 ± 0.008*||1.42 1,13.9||0.25|
Cell-mediated immune response
Neither common origin nor common rearing environment accounted for significant covariance in nestling cell-mediated immune response. Cell-mediated immune response was explained by natal female treatment in interaction with time and nestling age, and by rearing female treatment in interaction with nestling age (Table 3). Cell-mediated immune response increased with time elapsed between injection and time of assessment of response in nestlings hatched from eggs laid by carotenoid-supplemented females, while the immune response was independent of time among the control nestlings (Fig. 1). This suggests that carotenoid nestlings can still increase their immune response when nestlings from controls have reached a plateau, although nestlings from controls have higher early response. In addition, cell-mediated immune response increased with age in the control group, while it was independent of age in nestlings from the carotenoid group. That was the case for both natal and rearing female treatment, although the effect was less strong for natal female treatment (Fig. 2a,b). Eleven-day-old nestlings reared by carotenoid-fed females already show the same level of immune response as that reached by nestlings from controls at 15 days of age only. This suggests that the immune system is mature at an earlier age in nestlings reared by carotenoid-fed females. However, cell-mediated immune response of nestlings did not differ significantly according to natal female treatment (carotenoid: 0.72 ± 0.02 mm; control: 0.73 ± 0.02 mm; comparison of least square means: t = −0.83, d.f. = 132, P = 0.41) or rearing female treatment (carotenoid: 0.76 ± 0.02 mm; control: 0.70 ± 0.03 mm; comparison of least square means: t = 1.78, d.f. = 132, P = 0.08).
|Effect||Estimate ± SE||Fd.f.||P|
|Time between injection and control||−0.00003 ± 0.00008||5.43 1,132||0.02|
|Nestling age||0.08 ± 0.02||0.76 1,132||0.38|
|Natal female treatment||0.37 ± 0.29*||1.57 1,132||0.21|
|Rearing female treatment||1.07 ± 0.29*||13.19 1,132||0.0004|
|Time between injection and control × natal female treatment||0.0003 ± 0.0001*||8.07 1,132||0.005|
|Nestling age × natal female treatment||−0.06 ± 0.03*||4.96 1,132||0.03|
|Nestling age × rearing female treatment||−0.07 ± 0.02*||11.75 1,132||0.0008|
Nestlings from the same rearing nest were similar in total carotenoid concentration in plasma (covariance parameter estimate ± SE: 151.70 ± 60.28, Wald's Z = 2.52, P < 0.01), while that was not the case for nestlings from the same natal nest (covariance parameter estimate ± SE: 4.17 ± 24.68, Wald's Z = 0.17, P = 0.43) (likelihood ratio test: = 28.50, P < 0.0001). Nestling carotenoid concentration was not significantly affected by either natal (F[1,13.3] = 0.99, P = 0.34, carotenoid: 51.98 ± 2.16 μg mL−1, control: 50.49 ± 2.49 μg mL−1) or rearing female treatment (F[1,23.9] = 0.53, P = 0.48, carotenoid: 49.37 ± 2.17 μg mL−1, control: 53.40 ± 2.47 μg mL−1). Plasma carotenoid levels were therefore mostly influenced by rearing and environmental conditions, and not by genetic or carotenoid-mediated maternal effects.
This was also true for plasma vitamin E levels, whose variations were influenced by factors related to common characteristics of the pair of nests, rather than common origin or common-rearing environment. As total vitamin E concentration in nestling plasma is positively correlated with plasma carotenoids (r = 0.81, P < 0.0001), we investigated vitamin E in a model controlling for total carotenoid concentration. Total vitamin E concentration was not significantly affected by either natal (carotenoid: 12.27 ± 0.99 μg mL−1, control: 12.45 ± 1.12 μg mL−1) or rearing female treatment (carotenoid: 10.65 ± 0.84 μg mL−1, control: 14.32 ± 1.23 μg mL−1), but decreased with nestling age (Table 4).
|Covariance parameter estimates||Likelihood ratio test: = 13.90, P = 0.0002|
|Parameter||Estimate ± SE||Wald's Z||P|
|Pair||14.36 ± 8.49||1.69||0.04|
|Natal nest (Pair)||0.21 ± 1.32||0.16||0.44|
|Rearing nest (Pair)||2.19 ± 2.49||0.88||0.19|
|Residual||17.35 ± 2.22||7.81||<0.0001|
|Type 3 tests of fixed effects|
|Effect||Estimate ± SE||Fd.f.||P|
|Total carotenoid concentration in plasma||0.29 ± 0.02||294.92 1,148||<0.0001|
|Nestling age||−1.17 ± 0.48||5.91 1,20.6||0.02|
|Natal female treatment||−1.28 ± 0.70*||3.28 1,9.81||0.10|
|Rearing female treatment||−1.35 ± 0.97*||1.94 1,7.73||0.20|
As circulating carotenoids may affect pigment uptake by the follicle and pigment deposition in growing feathers, the effect of treatment on feather colour was analysed in relation to plasma carotenoid concentration.
Rearing conditions experienced during post-hatching development were a significant determinant of feather brightness in the yellow (rearing nest covariance parameter estimate ± SE: 60 975 ± 19 159, Wald's Z = 3.18, P < 0.001) while that was not the case for genetic and prehatching parental effects (natal nest covariance parameter estimate ± SE: 3889 ± 4141, Wald's Z = 0.94, P = 0.17) (likelihood ratio test: = 100.77, P < 0.0001). Feather brightness in the yellow was not affected by rearing female treatment (F[1,25] = 2.49, P = 0.13, carotenoid: 2744 ± 31, control: 2866 ± 29) but by the interaction between natal female treatment and total carotenoid concentration in plasma of nestlings (total carotenoid main effect: F[1,191] = 0.17, P = 0.68; natal treatment main effect: F[1,86.1] = 2.43, P = 0.12, carotenoid: 2820 ± 30; control: 2791 ± 31; interaction: F[1,165] = 4.35, P = 0.04, estimate = 2.98 ± 1.43). The positive estimate of the interaction term indicates that feather brightness increased more with plasma carotenoid levels in nestlings from eggs laid by carotenoid-supplemented females than in nestlings hatched from controls. The effect is, however, not strong enough to be clearly reflected in partial regressions (Fig. 3).
Similarly, variations in nestling feather colour hue were explained by common rearing environment but not by common origin, i.e. nestlings were more similar in hue within than among rearing nests (covariance parameter estimate ± SE: 88.19 ± 27.06, Wald's Z = 3.26, P < 0.001, natal nest covariance parameter not estimated, likelihood ratio test: = 102.97, P < 0.0001). Feather colour hue was not significantly affected by either natal (F[1,180] = 0.02, P = 0.89, carotenoid: 80.2 ± 1.1°, control: 80.5 ± 1.1°) or rearing female treatment (F[1,25.9] = 0.43, P = 0.52, carotenoid: 79.1 ± 1.1°, control: 81.57 ± 1.1°), or by nestling plasma carotenoid concentration (F[1,194] = 0.00, P = 0.95).
Maternal effects increase phenotypic plasticity in offspring traits and may therefore facilitate adaptation to environmental variability. However, assessing potential transgenerational and population consequences of maternal effects requires a better understanding of their mechanisms of action. In this experiment, we aimed at separating maternal effects through egg quality and through post-hatching maternal care that could both be mediated by antioxidant availability during egg laying. Pre- and post-hatching maternal effects were found to modulate independently or in concert nestling growth as reflected by wing length, immune function and plumage colour, but not body size and mass or plasma antioxidants (Table 5).
|Variable||Covariance parameter||Fixed effects|
|Nest||Maternal treatment||Interactions with covariates||Effect|
|Body mass||Natal, Rearing||–||–||–|
|Wing length||Natal, Rearing||Rearing × natal||Rearing × tarsus length||Not detectable|
|Sedimentation rate||Pair||Rearing||–||Lower in carotenoid nestlings compared to controls|
|Cell-mediated immune response||–||Rearing||Natal × time||Increase with time in carotenoid. No relationship with time in control|
|Natal × age||No relationship with age in carotenoid. Increase with age in control|
|Rearing × age||No relationship with age in carotenoid. Increase with age in control to reach in 15 d. old nestlings the same level as in carotenoid|
|Plasma vitamin E||Pair||–||–||–|
|Feather brightness||Rearing||–||Natal × plasma carotenoids||Increase more with plasma carotenoids in carotenoid|
|Feather colour hue||Rearing||–||–||–|
However, our experimental protocol has several limits that should be acknowledged. Cross-fostering hatchlings and not eggs did not allow us to control for any effect of carotenoid supplementation on female incubation capacity that is thus included in the natal treatment effects. As the sample size was relatively small, statistical power in estimating the relative importance of common origin and common rearing environment effects, and pre- and post-hatching maternal effects on nestling phenotypic traits may have been low. In addition, dietary supplementation of females during egg-laying probably resulted in a gradual increase in yolk carotenoid concentration with laying sequence (discussed in Biard et al., 2005). Thus, if the effect of egg carotenoids is concentration-dependent and moderate, then it might be detected only when the entire brood is considered, and cross-fostering broods and separating nestlings into two different rearing environments might have obscured this effect. Cross-fostered and noncross-fostered nests differed with respect to hatching date, as the distribution of hatching dates had lower variance and was skewed towards earlier dates in the former (see Material and methods). Thus, the seasonal effect was less important in cross-fostered nests, and under the assumption that hatching peak matches caterpillar abundance in this population (see Thessing & Ekman, 1994), cross-fostered nestlings have been reared during the optimal period of reproduction. The effects of female treatment might be more difficult to detect in such prime conditions than in poor conditions.
Indeed, although in noncross-fostered nests, nestlings from carotenoid-fed females achieved longer tarsi at fledging than nestlings from control females (Biard et al., 2005), in cross-fostered nestlings no effect of maternal effects through carotenoid availability could be detected on nestling body size or body mass. It was, however, possible to detect the influence of both pre- and post-hatching maternal effects on nestling growth, as measured by wing length, although we could not disentangle their interactive effects.
Nestlings reared by carotenoid-fed females had a lower blood sedimentation rate than nestlings of control females. Carotenoid-fed females may have been able to provide their nestlings with better growth conditions, in such a way that they may have been less subject to infections and inflammatory diseases than nestlings from controls. However, precise information on nestling infection status is not currently available. Cell-mediated immune response of cross-fostered nestlings showed a complex influence of both pre- and post-hatching maternal effects mediated by carotenoids. The immune response in nestlings from eggs laid by carotenoid-fed females was still increasing at the time of measure, when that of nestlings from controls did not increase with time, and probably had reached a plateau. In addition, our results suggest that the immune system of nestlings from carotenoid-fed females was mature at an earlier age than that of controls. This is consistent with noncross-fostered nestlings from carotenoid-fed females showing a more rapid development of the immune system and maturation of immune cells (Biard et al., 2005). Yolk carotenoids thus modulate the ontogeny and/or the dynamics of cell-mediated immune response (consistent with results of Saino et al., 2003), and the quality of post-hatching growth conditions provided by females is an additional important factor for the development of the immune system. Pre- and post-hatching maternal effects mediated by carotenoids may therefore have long-term consequences for the immune system of offspring and their capacity to resist pathogens in adulthood. We did not detect any significant covariance among the nestlings in components of immune function (i.e. sedimentation rate and cell-mediated immune response) according to common origin or common-rearing environment. However, previous cross-fostering experiments investigating the parasitism and immune function in passerines reported significant effects of common origin and common-rearing environments, their relative importance varying among the studies (Møller, 1990; Saino et al., 1997; Brinkhof et al., 1999; Soler et al., 2003; Cichoñet al., 2006). This suggests that common origin and common rearing effects previously interpreted as additive genetic variance and environmental effects may in fact have been due to pre- and post-hatching maternal effects. Cross-fostering designs are widely used in natural populations to estimate the relative contributions of genetic and environmental variances in the development of the phenotype. However, these designs rely on several assumptions (no dominance or epistasis genetic effects, negligible premanipulation parental effects) (Hoffmann & Merilä, 1999; Kruuk et al., 2001; Krist & Remeš, 2004) that are currently questioned especially for phenotypic traits closely associated with fitness (Crnokrak & Roff, 1995; Mousseau & Fox, 1998; Wolf et al., 2002; Pakkasmaa et al., 2003). These assumptions may lead to incorrect estimates of the relative importance of the different components of the phenotype. Associating a cross-fostering design with an experimental manipulation of maternal effects revealed that taking into account the specific influence of maternal effects mediated by carotenoids on the expression of immune function results in much lower covariance among the nestmates and siblings than previously estimated (see also Soler et al., 2003).
Plasma antioxidant levels were mostly influenced by rearing and environmental conditions, consistent with results from a cross-fostering study of North American kestrels Falco sparverius (Bortolotti et al., 2000). This reflects a stronger dependence of plasma carotenoid levels in nestlings on environmental or seasonal availability of relevant food items, linked to territory and/or parental quality, than on genetically based ability to metabolize the carotenoids. There was no effect of female treatment on circulating levels of carotenoids or vitamin E in nestlings, as in noncross-fostered nestlings (Biard et al., 2005). Plasma vitamin E was positively related to plasma carotenoid levels, which may reflect the interaction between antioxidants, or the possibility that some carotenoid dietary sources are also rich in vitamin E, therefore providing both antioxidants simultaneously. Plasma vitamin E decreased with nestling age, probably reflecting the depletion of yolk-derived reserves during the ontogeny (Surai, 2002).
The development of juvenile carotenoid-based plumage colour was determined by common-rearing environment, but not by common origin, which contrasts with previous studies of the closely related great tit Parus major showing an influence of both common origin and common-rearing environments on plumage colour (Slagsvold & Lifjeld, 1985; Hõrak et al., 2000; Fitze et al., 2003; Tschirren et al., 2003). Natal female treatment and thus egg quality influenced plumage colour in a slightly different way than in noncross-fostered nests, where nestlings from carotenoid-supplemented females produced brighter yellow feathers than nestlings from controls. In cross-fostered nestlings, feather brightness increased more with plasma carotenoid levels in nestlings from eggs laid by carotenoid-supplemented females than in nestlings hatched from controls. This indicates that yolk carotenoid concentration may have a long-term influence on the ability of nestlings to incorporate plasma carotenoids into growing feathers, with potential consequences in adulthood on the development of sexual signals. It is indeed known that availability of yolk-derived carotenoids during early developmental stages influences the subsequent absorption and utilization of dietary carotenoids during post-natal life (Koutsos et al., 2003). An impaired ability to metabolize and use the carotenoids may also have consequences for the immune and antioxidant systems. Moreover, a previous experiment in our population showed that dietary supplementation of blue tit nestlings with carotenoids during growth did not enhance plumage colour, while that was the case for great tit nestlings. This interspecific difference was hypothesized to be a consequence of blue tit nestlings originating from the eggs with lower carotenoid content and developing faster than great tit nestlings (Biard et al., 2006). The results presented here confirm that the ability to absorb and/or use the carotenoids and consequently the development of carotenoid-based plumage in blue tit nestlings is affected by the amount of carotenoids deposited in egg yolk. These results also suggest that common origin effects found in previous studies of tits and interpreted as genetic differences (e.g. Fitze et al., 2003) may be a consequence of maternal effects mediated by female investment in eggs. In contrast, hue of nestling feathers was not influenced by female treatment in cross- or noncross-fostered nestlings. This suggests that the relationship among the yellow plumage colour, the carotenoid content of plumage and its structural quality in relation to carotenoid status of the individual needs to be further investigated in Parus species (e.g. Saks et al., 2003a; Shawkey & Hill, 2005).
In this experiment, we have shown that carotenoid availability influenced the development of nestling phenotype, and that nestling characteristics were modulated by yolk carotenoids, post-hatching maternal effects, or both. Yolk carotenoids could affect nestling development through their role in regulating the oxidative stress produced by growth. In addition, egg composition could also modulate gene expression during embryonic development (Zhang et al., 1992; Bertram & Borthiewicz, 1995; Burri, 2000) potentially affecting various metabolic functions in post-natal life such as enzymatic systems responsible for assimilation and metabolism of natural antioxidants, including carotenoids (Surai, 2002). How could the availability of carotenoids during egg laying influence the quality of post-hatching growth conditions provided to nestlings by females? At the time of chick rearing, carotenoid-supplemented females had increased plasma vitamin E levels compared with the controls (Biard et al., 2005). As a consequence, carotenoid-supplemented females may have been less in need of antioxidant-rich food than control females, and this may have resulted in a relative increase in antioxidant-rich items in food delivered to nestlings. Females from the two treatment groups may also have differed in other brood-rearing behaviours, like nest cleaning or brooding at night, and this may have provided nestlings with growth environments of different quality. The availability of carotenoids to egg laying females thus affected their capacity to lay high quality eggs, but also their ability to provide chicks with optimal growth conditions. Females may therefore have to balance the benefits of investing in egg quality against the potential costs of impairing their own future antioxidant protection, not only because this may have consequences for their survival, but also because this may decrease their brood-rearing capacity and thereby negatively affect offspring fitness.
Maternal effects mediated by carotenoids may modulate the expression of components of the phenotype that are closely associated with nestling fitness, such as immune response and ability to metabolize and use antioxidants, through direct effects of yolk antioxidants and by modifying early developmental conditions and the growth environment of nestlings. Maternal effects on traits such as immune response and antioxidant function may not only increase offspring fitness, but also have important consequences at the population level. Immune response and antioxidant function are likely to be under strong directional selection, and beneficial alleles affecting these functions are likely to go to fixation (Christe et al., 2000; Endler, 2000; Kruuk et al., 2001), decreasing adaptation capacity to changes in the environment (i.e. prevalence of parasites or carotenoid availability). However, maternal effects may under certain circumstances facilitate adaptation to the environment. If maternal effects act in the same direction as that of natural selection, maternal effects may speed up the rate of evolution without depleting the genetic variation (Wade, 1998). Maternal effects mediated by antioxidants may be purely environmental, if relying only on carotenoid availability in the diet, or females may also differ in their ability to acquire and use carotenoids as a result of genetic variability in physiological processes related to absorption, assimilation and use (Olson & Owens, 1998). Indirect genetic effects may help maintain genetic variation in immune response because of increase in phenotypic plasticity and epistatic interactions between maternal and offspring genotypes (Wade, 1998) and provide additional variation that may facilitate phenotypic evolution (evolving environment Wolf et al., 1998). Maternal effects mediated by antioxidants and particularly carotenoids may thus act as a proximate factor in development and phenotypic plasticity, but also ultimately in the evolution of phenotypic traits.
Bird capture, ringing, supplemental feeding, cross-fostering and blood sampling were performed under the Ringing Licence (CRBPO Museum National d'Histoire Naturelle) and special authorization for protected species from Direction Régionale de l'Environnement (DIREN Champagne-Ardenne). We gratefully acknowledge the help of Parc Naturel Regional de la Forêt d'Orient and Office National des Forêts. Kemin Europa provided samples of carotenoids (OroGlo Layer Dry). We thank B. Auclair, S. Cassier and V. Coutouly for help in the field. G.E. Hill, J. Merilä and T. Rigaud provided helpful comments on previous versions of the manuscript. D. Fairbairn and two reviewers provided suggestions that greatly improved the manuscript. C.B. was supported by a doctoral grant from Ministère de l'Education et de la Recherche and P.F.S. was supported by the Scottish Executive Environment and Rural Affairs Department.
- 1993. Physiological roles of antioxidants in the immune system. J. Dairy Sci. 76: 2789–2794.
- 1995. Dietary carotenoids inhibit neoplastic transformation and modulate gene expression in mouse and human cells. Am. J. Clin. Nutr. 62: 1327–1336. &
- 2005. Effects of carotenoid availability during laying on reproduction in the blue tit. Oecologia 144: 32–44. , &
- 2006. Carotenoid availability in diet and phenotype of blue and great tit nestlings. J. Exp. Biol. 209: 1004–1015. , &
- 2000. Why egg yolk is yellow. Trends Ecol. Evol. 15: 47–49. , &
- 2002. Carotenoids and egg quality in the lesser black-backed gull (Larus fuscus): a supplemental feeding study of maternal effects. Proc. R. Soc. Lond. B 269: 29–36. , , , , , &
- 2003. Neonatal nutrition, adult antioxidant defences and sexual attractiveness in the zebra finch. Proc. R. Soc. Lond. B 270: 1691–1696. , , , , &
- 2004. Egg-laying capacity is limited by carotenoid pigment availability in wild gulls Larus fuscus. Proc. R. Soc. Lond. B 271 (Suppl.): S79–S81. , , &
- 2000. Genetics, local environment and health as factors influencing plasma carotenoids in wild American kestrels (Falco sparverius). Proc. R. Soc. Lond. B 267: 1433–1438. , , , &
- 1999. Immunocompetence of nestling great tits in relation to rearing environment and parentage. Proc. R. Soc. Lond. B 266: 2315–2322. , , &
- 1998. Model Selection and Inference: A Practical Information-Theoretic Approach. Springer Verlag, New York. &
- 2000. Carotenoids and gene expression. Nutrition 16: 577–578.
- 1993. Role of carotenoids in the immune response. J. Dairy Sci. 76: 2804–2811.
- 2000. Genetic and environmental components of phenotypic variation in immune response and body size of a colonial bird, Delichon urbica (the house martin). Heredity 85: 75–83. , , &
- 2006. Genetic and environmental variation in immune response of collared flycatcher nestlings. J. Evol. Biol. 19: 1669–1674. , &
- 1995. Dominance variance: associations with selection and fitness. Heredity 75: 530–540. &
- 1996. Poultry Immunology. Carfax Publishing Company, Oxford. , &
- 1990. On the measurement and classification of colour in studies of animal colour patterns. Biol. J. Linn. Soc. 41: 315–352.
- 2000. Adaptive genetic variation in the wild. In: Adaptive Genetic Variation in the Wild (T. A.Mousseau, B.Sinervo & J. A.Endler, eds), pp. 251–260. Oxford University Press, Oxford.
- 1996. Introduction to Quantitative Genetics, 4th edn. Longman, New York. ,
- 2003. Effects of common origin and common environment on nestling plumage coloration in the great tit (Parus major). Evolution 57: 144–150. , &
- 2002. On the misuse of residuals in ecology: regression of residuals vs. multiple regression. J. Anim. Ecol. 71: 542–545.
- 1991. Nestling growth in the great tit. I. Heritability estimates under different environmental conditions. J. Evol. Biol. 4: 341–362. &
- 2003. Multilevel Statistical Models, 3rd edn. Hodder Arnold, London.
- 1984. The Biochemistry of Carotenoids. Chapman and Hall, London.
- 1997. Parental preference for red mouth of chicks in a songbird. Proc. R. Soc. Lond. B 264: 959–962. &
- 2002. Dietary carotenoids predict plumage coloration in wild house finches. Proc. R. Soc. Lond. B 269: 1119–1124. , &
- 1999. Heritable variation and evolution under favourable and unfavourable conditions. Trends Ecol. Evol. 14: 96–101. &
- 2000. Growth conditions affect carotenoid-based plumage coloration of great tit nestlings. Naturwissenschaften 87: 460–464. , , &
- 2002. Fat-soluble antioxidants in the eggs of great tits Parus major in relation to breeding habitat and laying sequence. Avian Sci. 2: 123–130. , &
- 2003. Maternal carotenoid status modifies the incorporation of dietary carotenoids into immune tissues of growing chickens (Gallus gallus domesticus). J. Nutr. 133: 1132–1138. , , &
- 2004. Maternal effects and offspring performance: in search of the best method. Oikos 106: 422–426. &
- 2001. Phenotypic selection on a heritable size trait revisited. Am. Nat. 158: 557–571. , &
- 2003. Fluctuating asymmetry in a secondary sexual trait: no associations with individual fitness, environmental stress or inbreeding, and no heritability. J. Evol. Biol. 16: 101–113. , , &
- 1987. Unrepeatable repetabilities: a common mistake. Auk 104: 116–121. &
- 1990. Effects of a haematophagous mite on the barn swallow (Hirundo rustica): a test of the Hamilton and Zuk hypothesis. Evolution 44: 771–784.
- 2000. Carotenoid-dependant signals: indicators of foraging efficiency, immunocompetence or detoxification ability? Avian Poultry Biol. Rev. 11: 137–159. , , , , , &
- 2005. Maternally derived carotenoid pigments affect offspring survival, sex-ratio, and sexual attractiveness in a colorful songbird. Naturwissenschaften 92: 375–380. , &
- 1996. Genetic variation in offspring condition: an experiment. Funct. Ecol. 10: 465–474.
- 1997. Expression of genetic variation in body size of the collared flycatcher under different environmental conditions. Evolution 51: 526–536.
- 1999. Genetic variation and natural selection on blue tit body condition in different environments. Genet. Res. 73: 165–176. , &
- 1998. Maternal Effects as Adaptations. pp. 375. Oxford University Press, Oxford. &
- 1987. Natural selection and the heritability of fitness components. Heredity 59: 181–198. &
- 2001. Differential post-fledging survival of great and coal tits in relation to their condition and fledging date. J. Anim. Ecol. 70: 730–738. , &
- 1998. Costly sexual signals: are carotenoids rare, risky or required? Trends Ecol. Evol. 13: 510–514. &
- 2003. Genetic and maternal effect influences on viability of common frog tadpoles under different environmental conditions. Heredity 91: 117–124. , &
- 1998. Kinetic and immunohistochemical characteristics of mitogen-induced cutaneous hypersensitivity in chickens selected for antibody responsiveness. Vet. Immunol. Immunopathol. 66: 367–376. , &
- 1987. Carotenoids in food chain studies. II. The food chain of Parus spp. Monitored by carotenoid analysis. Comp. Biochem. Physiol. B 87: 885–888. , , &
- 1998. Maternal and paternal effects in birds. Effects on offspring fitness. In: Maternal Effects as Adaptations (T. A.Mousseau & C. W.Fox, eds), pp. 202–226. Oxford University Press, Oxford.
- 1991. On the low heritability of life-history traits. Evolution 45: 853–861. &
- 1998. The detection and measurement of maternal effects. In: Maternal Effects as Adaptations (T. A.Mousseau & C. W.Fox, eds), pp. 83–96. Oxford University Press, Oxford.
- 2001. Maternally derived androgens and antioxidants in bird eggs: complementary but opposing effects? Behav. Ecol. 12: 381–385. , &
- 2003. The effect of variation in dietary intake on maternal deposition of antioxidants in zebra finch eggs. Funct. Ecol. 17: 472–481. , &
- 1997. Immunocompetence of nestling barn swallows in relation to brood size and parental effort. J. Anim. Ecol. 66: 827–836. , &
- 1999. Carotenoid plasma concentration, immune profile and plumage ornamentation of male barn swallows (Hirundo rustica). Am. Nat. 154: 441–448. , , , &
- 2002. Carotenoid concentration in barn swallow eggs is influenced by laying order, maternal infection and paternal ornamentation. Proc. R. Soc. Lond. B 269: 1729–1733. , , , , &
- 2003. Experimental manipulation of egg carotenoids affects immunity of barn swallow nestlings. Proc. R. Soc. Lond. B 270: 2485–2489. , , , &
- 2003a. How feather colour reflect its carotenoid content. Funct. Ecol. 17: 555–561. , &
- 2003b. Carotenoid-based plumage coloration of male greenfinches reflects health and immunocompetence. Oecologia 134: 301–307. , &
- 1990. Avian Cellular Immunology. CRC Press, Boca Raton.
- 2005. Carotenoids need structural colours to shine. Biol. Let. 1: 121–124. &
- 1985. Variation in plumage colour of the great tit Parus major in relation to habitat, season and food. J. Zool. 206: 321–328. &
- 1995. Heritability of nestling growth in cross-fostered European starlings Sturnus vulgaris. Genetics 141: 657–665. &
- 1999. Simplifying the phytohaemagglutinin skin-testing technique in studies of avian immunocompetence. Funct. Ecol. 13: 567–572. , &
- 2003. Environmental, genetic and maternal components of immunocompetence of nestling pied flycatchers from a cross-fostering study. Evol. Ecol. Res. 5: 259–272. , &
- 2005. A simulation study on tests of hypotheses and confidence intervals for fixed effects in mixed models for block experiments with missing data. J. Agr. Biol. Envir. St. 10: 374–389. , &
- 1986. Avian Physiology, 4th edn. Springer-Verlag, New York.
- 2000. Effect of selenium and vitamin E content of the maternal diet on the antioxidant system of the yolk and the developing chick. Brit. Poultry Sci. 41: 235–243.
- 2002. Natural Antioxidants in Avian Nutrition and Reproduction. Nottingham university Press, Nottingham, UK.
- 2001a. Carotenoids in avian nutrition and embryonic development. 2. Antioxidant properties and discrimination in embryonic tissues. J. Poultry Sci. 38: 117–145. , &
- 2001b. Carotenoids in avian nutrition and embryonic development. 1. Absorption, availability and levels in plasma and egg yolk. J. Poultry Sci. 38: 1–27. , &
- 2001c. Carotenoid discrimination by the avian embryo: a lesson from wild birds. Comp. Biochem. Physiol. B 128: 743–750. , , , , &
- 1998. Molt and migratory condition in blue tits: a serological study. Condor 98: 825–831. &
- 1994. Selection on the genetical and environmental components of tarsal growth in juvenile willow tits. J. Evol. Biol. 7: 713–726. &
- 1990. Nestling weight and survival in individual great tits (Parus major). J. Anim. Ecol. 59: 1113–1127. &
- 2003. Proximate mechanisms of variation in the carotenoid-based plumage coloration of nestling great tits (Parus major L.). J. Evol. Biol. 16: 91–100. , &
- 1998. The evolutionary genetics of maternal effects. In: Maternal Effects as Adaptations (T. A.Mousseau & C. W.Fox, eds), pp. 5–21. Oxford University Press, Oxford.
- 1998. Evolutionary consequences of indirect genetic effects. Trends Ecol. Evol. 13: 64–69. , , , &
- 2002. Contribution of maternal effect QTL to genetic architecture of early growth in mice. Heredity 89: 300–310. , , &
- 1992. Carotenoids up-regulate connexin 43 gene expression independent of pro-vitamin A or antioxidant properties. Cancer Res. 52: 5707–5712. , &